This article examines the extent to which self-reported emotional eating is a predictor of unhealthy snack consumption or, alternatively, an expression of beliefs about the relation between emotions and eating derived from concerns about eating behaviour. Three studies were conducted. Study 1 (N = 151) and Study 2 (N = 184) investigated the predictive validity of emotional eating compared to habit strength in snack consumption, employing 7-day snack diaries. Both studies demonstrated that snack consumption was not predicted by emotional eating but depended on the habit of unhealthy snacking and on restraint eating. As emotional eating was not a significant predictor of snack intake, Study 3 addressed the alternative hypothesis of emotional eating being an expression of concerns about eating behaviour. Results from this cross-sectional survey (N = 134) showed that emotional eating was significantly associated with several concerns. Together, these studies show that snack intake is better predicted by habit strength and restraint eating than by emotional eating. Additionally, the results suggest that in normal-weight women the concept of emotional eating may not capture the tendency to eat under emotional conditions, but rather reflects beliefs about the relation between emotions and eating.